The goals of this proposal are to use cytogenetic and molecular biological approaches to define the genetic mechanisms of childhood brain tumors and to determine the biological significance of these findings as they relate to the initiation, etiology, progression and responsiveness to therapy of CNS tumors. Preliminary data suggests that individual types of childhood brain tumors possess specific cytogenetic and molecular genetic characteristics which are different from those seen in adult nervous system tumors. Theses studies will determine the chromosomal abnormalities of childhood brain tumors, establish cultured cell lines and transplantable xenografts which maintain these karyotypic abnormalities and use these models to determine which chromosomal regions and individual genes are altered by amplification, translocation, or deletion. The specific aims are: 1) To define the specific chromosomal abnormalities that characterize individual types of childhood brain tumors including medulloblastomas, brainstem gliomas, ependymomas, supratentorial gliomas, and cerebellar astrocytomas in order to determine if these tumors have karyotypic profiles which are different from one another and from neuroblastomas and adult gliomas. 2) To determine the incidence of gene amplification and the spectrum of genes which are amplified in these individual types of childhood tumors. 3) To establish a bank of frozen, karyotyped childhood brain tumors and permanent cell lines and xenografts established from them to allow molecular characterization of the chromosomal abnormalities. Specifically restriction fragment length polymorphisms and in situ hybridization will be used to confirm numerical deviations and translocations and to define regions of chromosomal deletions. These studies will define the cytogenetic and molecular mechanisms which are important in the initiation and progression of childhood brain tumors and will identify subgroups of patients with similar cytogenetic or molecular genetic abnormalities whose tumors may progress more rapidly or may respond differently to specific types of therapeutic intervention. Specifically, these studies will determine if gene amplification is indicative of rapid tumor progression as has been shown for neuroblastoma. In addition, regions of chromosomes which are consistently deleted will be defined cytogenetically and molecular genetically providing the basis for the isolation and characterization of genes whose loss are central to the initiation or progression of these tumors.